Process Capability and

Statistical Quality Control

Topics
 Introduction
 Process variations
 Assignable vs. common
 Specifications/ tolerance limits
 Process Capability
 Capability index
 Statistical Process Control (SPC) procedures
 Attribute measurements – p charts
 Variable measurements – x bar & R charts
 Acceptance Sampling
 Single sampling plan for attributes
 Operating Characteristic (OC) Curves
Basic Forms of Variation
Assignable variation is caused
by factors that can be Example: A poorly trained
employee that creates variation
clearly identified and in finished product output.
possibly managed

Common variation is
Example: A molding process that
inherent in the production always leaves “burrs” or flaws on
process a molded item.
 Taguchi’s View of Variation
Traditional view is that quality within the LSL & USL is good and that the cost
of quality outside this range is constant, whereas Taguchi views costs as
increasing as variability increases, so seek to achieve zero defects and that will
truly minimize quality costs.

High High

Incremental Incremental
Cost of Cost of
Variability Variability

Zero Zero

Lower Target Upper Lower Target Upper

Spec Spec Spec Spec Spec Spec

Traditional View Taguchi’s View

Process Capability

 Tolerances
 Choice of process & its consistency
 Perfectly centered process
 Process limits
 Specification limits
 How do the limits relate to one another?
 Process capability
 How well the process is capable of producing parts relative to the
design tolerances
Process Capability Index, Cpk
Capability Index shows how  X  LTL UTL - X 
well parts being produced fit C pk = min  or 
into the range specified by the  3 3 
design limits

As a production process
produces items small shifts
in equipment or systems
can cause differences in
production performance
from differing samples.

Shifts in Process Mean

Process Capability – A Standard
Measure of How Good a Process Is

A simple ratio:
 Specification width/ Actual process width
 Generally, the bigger the better
Process Capability

 X  LTL UTL  X 
C pk  Min  ; 
 3 3 

This is a “one-sided” Capability Index

Concentration on the side which is closest to the
specification - closest to being “bad”
The Cereal Box Example
 We are the maker of this cereal. Consumer reports has just
published an article that shows that we frequently have less
than 15 ounces of cereal in a box.
 Let’s assume that the government says that we must be within
± 5 percent of the weight advertised on the box (16 ounce).
 Upper Tolerance Limit = 16 + .05(16) = 16.8 ounces
 Lower Tolerance Limit = 16 – .05(16) = 15.2 ounces
 We go out and buy 1,000 boxes of cereal and find that they
weight an average of 15.875 ounces with a standard deviation
of .529 ounces.
Cereal Box Process Capability

 Specification or
Tolerance Limits  X  LTL UTL  X 
Upper Spec = 16.8 oz C pk  Min  ; 
 3 3 

 Lower Spec = 15.2 oz
 Observed Weight
 Mean = 15.875 oz 15.875  15.2 16.8  15.875 
C pk  Min  ; 
 Std Dev = .529 oz
 3(.529) 3(.529) 

C pk  Min.4253; .5829

C pk  .4253
What does a Cpk of .4253 mean?

 An index that shows how well the units being produced

fit within the specification limits.
 This is a process that will produce a relatively high
number of defects.
 Many companies look for a Cpk of 1.3 or better… 6-
Sigma company wants 2.0!
 Process
Lower Tolerance Mean = 15.875 Upper Tolerance
= 15.2 Std. Dev. = .529 = 16.8

What percentage of boxes are defective (i.e. less than 15.2 oz)?
Z LTL= (x – Mean)/Std. Dev. = (15.2 – 15.875)/.529 = -1.276
NORMSDIST(Z) = NORMSDIST(-1.276) = .100978
Approx. 10 % of the boxes have less than 15.2 ounces of cereal in them!

Similarly,
Z UTL= (x – Mean)/Std. Dev. = (16.8 – 15.875)/.529 = 1.748
NORMSDIST(Z) = NORMSDIST(1.748) = .0402
Approx. 4 % of the boxes have more than 16.8 ounces of cereal in them!
Process Control Procedures
 Process Control
 Monitoring quality when the product is being produced
 Provide timely information on whether products are meeting
design specifications
 process has shifted
 Statistical Process Control
 Testing random sample from a process
 Is the process producing items within a preselected range?
 Types of quality characteristics
 Attribute (Go or no-go information)
 Defectives refers to the acceptability of product across a
range of characteristics
 Defects refers to the number of defects per unit which may
be higher than the number of defectives
 p-chart application
 Variable (Continuous)
 Usually measured by the mean and the standard deviation
 X-bar and R chart applications
Statistical UCL

Process Normal Behavior

Control
(SPC) Charts LCL

1 2 3 4 5 Samples
6 over time
UCL

Possible problem, investigate

LCL

1 2 3 4 5 Samples
6 over time
UCL

Possible problem, investigate

LCL

1 2 3 4 5 Samples
6 over time
Control Limits based on the Normal Curve

x
m
z
-3 -2 -1 0 1 2 3
Standard
deviation units or
“z” units.
Control Limits
We establish the Upper Control Limits (UCL) and
the Lower Control Limits (LCL) with plus or minus 3
standard deviations from some x-bar or mean value.
Based on this we can expect 99.7% of our sample
observations to fall within these limits.

99.7%
x
LCL UCL
Example of Constructing a p-Chart:
Required Data

Sample No. of Number of

defects found in
No. Samples each sample
1 100 4
2 100 2
3 100 5
4 100 3
5 100 6
6 100 4
7 100 3
8 100 7
9 100 1
10 100 2
11 100 3
12 100 2
13 100 2
14 100 8
15 100 3
Statistical Process Control Formulas:
Attribute Measurements (p-Chart)

Given: T o ta l N u m b e r o f D e fe c tiv e s
p =
T o ta l N u m b e r o f O b s e rv a tio n s

p (1 - p)
sp =
n
Compute control limits:

UCL = p + z sp
LCL = p - z sp
Constructing a p-chart: Step 1

Sample n Defectives p
1. Calculate the 1 100 4 0.04
sample proportions, 2 100 2 0.02
3 100 5 0.05
p (these are what 4 100 3 0.03
can be plotted on the 5 100 6 0.06
p-chart) for each 6 100 4 0.04
7 100 3 0.03
sample 8 100 7 0.07
9 100 1 0.01
10 100 2 0.02
11 100 3 0.03
12 100 2 0.02
13 100 2 0.02
14 100 8 0.08
15 100 3 0.03
Constructing a p-chart: Steps 2&3

2. Calculate the average of the sample proportions

55
p = = 0.036
1500
3. Calculate the standard deviation of the
sample proportion

p (1 - p) .036(1- .036)
sp = = = .0188
n 100
 Constructing a p-chart: Step 4

4. Calculate the control limits

UCL = p + z sp
LCL = p - z sp

.036  3(.0188)

UCL = 0.0924
LCL = -0.0204 (or 0)
Constructing a p-Chart: Step 5

5. Plot the individual sample proportions, the average

of the proportions, and the control limits

0.16

0.14

0.12

0.1 UCL
p 0.08

0.06

0.04

0.02

0 LCL
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15

O b servation
x-bar and R Charts: Required Data

Sample Obs 1 Obs 2 Obs 3 Obs 4 Obs 5

1 10.68 10.689 10.776 10.798 10.714
2 10.79 10.86 10.601 10.746 10.779
3 10.78 10.667 10.838 10.785 10.723
4 10.59 10.727 10.812 10.775 10.73
5 10.69 10.708 10.79 10.758 10.671
6 10.75 10.714 10.738 10.719 10.606
7 10.79 10.713 10.689 10.877 10.603
8 10.74 10.779 10.11 10.737 10.75
9 10.77 10.773 10.641 10.644 10.725
10 10.72 10.671 10.708 10.85 10.712
11 10.79 10.821 10.764 10.658 10.708
12 10.62 10.802 10.818 10.872 10.727
13 10.66 10.822 10.893 10.544 10.75
14 10.81 10.749 10.859 10.801 10.701
15 10.66 10.681 10.644 10.747 10.728
x-bar and R charts: Step 1
Calculate sample means, sample ranges, mean of means, and
mean of ranges.
Sample Obs 1 Obs 2 Obs 3 Obs 4 Obs 5 Avg Range
1 10.68 10.689 10.776 10.798 10.714 10.732 0.116
2 10.79 10.86 10.601 10.746 10.779 10.755 0.259
3 10.78 10.667 10.838 10.785 10.723 10.759 0.171
4 10.59 10.727 10.812 10.775 10.73 10.727 0.221
5 10.69 10.708 10.79 10.758 10.671 10.724 0.119
6 10.75 10.714 10.738 10.719 10.606 10.705 0.143
7 10.79 10.713 10.689 10.877 10.603 10.735 0.274
8 10.74 10.779 10.11 10.737 10.75 10.624 0.669
9 10.77 10.773 10.641 10.644 10.725 10.710 0.132
10 10.72 10.671 10.708 10.85 10.712 10.732 0.179
11 10.79 10.821 10.764 10.658 10.708 10.748 0.163
12 10.62 10.802 10.818 10.872 10.727 10.768 0.250
13 10.66 10.822 10.893 10.544 10.75 10.733 0.349
14 10.81 10.749 10.859 10.801 10.701 10.783 0.158
15 10.66 10.681 10.644 10.747 10.728 10.692 0.103

Averages 10.728 0.220400

x-bar and R charts: Step 2
Determine Control Limit Formulas and Necessary Tabled Values

x Chart Control Limits

n A2 D3 D4
UCL = x + A 2 R 2 1.88 0 3.27
3 1.02 0 2.57
LCL = x - A 2 R 4 0.73 0 2.28
5 0.58 0 2.11
6 0.48 0 2.00
R Chart Control Limits 7 0.42 0.08 1.92
8 0.37 0.14 1.86
UCL = D 4 R 9 0.34 0.18 1.82
10 0.31 0.22 1.78
LCL = D 3 R 11 0.29 0.26 1.74
 x-bar and R charts: Steps 3&4
Calculate x-bar Chart and Plot Values

UCL = x + A 2 R  10.728 - .58(0.2204 ) = 10.856

LCL = x - A 2 R  10.728 - .58(0.2204 ) = 10.601
1 0 .9 0 0

1 0 .8 5 0 UCL
1 0 .8 0 0

1 0 .7 5 0
M ea n s

1 0 .7 0 0

1 0 .6 5 0

1 0 .6 0 0
LCL
1 0 .5 5 0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
S am p le
x-bar and R charts: Steps 5&6
Calculate R-chart and Plot Values
UCL = D 4 R  ( 2.11)(0.2204)  0.46504
LCL = D3 R  (0)(0.2204)  0

0 .8 0 0

0 .7 0 0

0 .6 0 0

0 .5 0 0
UCL
R 0 .4 0 0

0 .3 0 0

0 .2 0 0

0 .1 0 0

0 .0 0 0
LCL
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
S a m p le
Basic Forms of Statistical Sampling for
Quality Control
 Acceptance Sampling
 sampling to accept or reject the immediate lot of
product at hand
 Percentage of products conforming to specifications
 Statistical Process Control
 sampling to determine if the process is within acceptable
limits
Acceptance Sampling
 Purposes
 Determine quality level
 Ensure quality is within predetermined level

 Advantages
 Economy
 Less handling damage
 Fewer inspectors
 Upgrading of the inspection job
 Applicability to destructive testing
 Entire lot rejection (motivation for improvement)
Acceptance Sampling
 Disadvantages
 Risks of accepting “bad” lots and rejecting “good” lots
 Added planning and documentation
 Sample provides less information than 100-percent inspection
Acceptance Sampling:
Single Sampling Plan
 Decision about lot from one sample of items selected
at random from a lot

 Determine
1. how many units, n, to sample from a lot, and
2. the maximum number of defective items, c, that can be
found in the sample before the lot is rejected
Risk
 Acceptable Quality Level (AQL)
 Max. acceptable percentage of defectives defined by
producer
 The α (Producer’s risk)
 The probability of rejecting a good lot
 Lot Tolerance Percent Defective (LTPD)
 Percentage of defectives that defines consumer’s
rejection point
 The  (Consumer’s risk)
 The probability of accepting a bad lot
 Operating Characteristic Curve
The OCC brings the concepts of producer’s risk, consumer’s risk, sample size,
and maximum defects allowed together

1 The shape or
0.9 a = .05 (producer’s risk) slope of the
0.8 curve is
dependent on
Probability of acceptance

0.7 n = 99 a particular
0.6 c=4 combination of
0.5 the four
0.4 parameters
0.3  =.10
0.2 (consumer’s risk)
0.1
0
1 2 3 4 5 6 7 8 9 10 11 12
AQL LTPD
Percent defective
Example: Acceptance Sampling Problem
Zypercom, a manufacturer of video interfaces, purchases
printed wiring boards from an outside vender, Procard.
Procard has set an acceptable quality level of 1% and
accepts a 5% risk of rejecting lots at or below this level.
Zypercom considers lots with 3% defectives to be
unacceptable and will assume a 10% risk of accepting a
defective lot.

Develop a sampling plan for Zypercom and determine a

rule to be followed by the receiving inspection personnel.
Step 1. What is given and what is not?

In this problem, AQL is given to be 0.01 and LTDP

is given to be 0.03. We are also given an alpha of
0.05 and a beta of 0.10.

What you need to determine in your sampling

plan is “c” and “n.”
 Step 2. Determine “c”
LTPD .03
= = 3
First divide LTPD by AQL. AQL .01

Then find the value for “c” by selecting the value in the
TN8.10 “n(AQL)”column that is equal to or just greater than
the ratio above.

Exhibit TN 8.10 So, c = 6.

c LTPD/AQL n AQL c LTPD/AQL n AQL

0 44.890 0.052 5 3.549 2.613
1 10.946 0.355 6 3.206 3.286
2 6.509 0.818 7 2.957 3.981
3 4.890 1.366 8 2.768 4.695
4 4.057 1.970 9 2.618 5.426
Step 3. Determine Sample Size
Now given the information below, compute the sample size in units to
generate your sampling plan

c = 6, from Table
n (AQL) = 3.286, from Table
AQL = .01, given in problem

n(AQL/AQL) = 3.286/.01 = 328.6, or 329 (always round up)

Sampling Plan:
Take a random sample of 329 units from a lot.
Reject the lot if more than 6 units are defective.